Distributed detection and isolation of bias injection attack in smart energy grid via interval observer

With the integration in information and communication technologies, and advanced metering infrastructure, smart energy grid, as one of typical sustainable energy systems, addresses the energy and environment problems. However, the emergency of bias injection attack aiming at destroying the energy management center, brings great security threat to the security of smart energy grid. To address risks in energy-cyber-physical systems, this paper proposes a distributed detection and isolation scheme against the bias injection attack in smart energy grid. Considering the transmitted information of energy management centers in adjacent grid subareas, the proposed distributed detection and isolation scheme includes local and global steps. In the local-step, each local energy management center detects and isolates the possible sensor attack set, based on the constructed local attack signature judgment logic matrix. In the global-step, the subarea attack set is detected and isolated via the established global attack signature judgment logic matrix. Combining the above local and global detection and isolation framework, we can ensure the security of energy management center in smart energy system. This proposed distributed detection and isolation scheme examines some important practical aspects of deploying bias injection attack detection including: the limitation of the precomputed threshold; the detection delay; the accuracy in detecting bias injection attack. Finally, the effectiveness of the developed distributed detection and isolation scheme is demonstrated by using detailed studies on the IEEE 8-bus and IEEE 118-bus smart energy grid system.